rdata.table

How to apply vector function on data.table

A function to sum 2 vectors

vsum = \(s) s[1] + s[2]

works well for data.frame

> vsum(expand.grid(1:2,2:3))
  Var1
1    3
2    4
3    4
4    5

but how to get the same result for data.table?

This is incorrect result:

> library(data.table)
> vsum(CJ(1:2,2:3))
      V1    V2
   <int> <int>
1:     2     5

Additional notes

I do not want the function with two variables.

Solution

  • Changing the function to

    vsum = \(s){ s[,1,drop=F] + s[,2,drop=F] }

    should give the desired result for both, data.frame and data.table.

    Choosing columns with s[1] is a data.frame shortcut. With data.table it picks the row.